Closed-Volume Leak Detection Logic Using Multi-Sensor Cross-Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing leak detection methods in closed-volume liquid systems, such as those used in aircraft cooling systems, are not always reliable and may provide inaccurate readings, necessitating the need for improved logic to quickly detect and isolate leaks before significant fluid loss occurs, which can lead to system failure, fire hazards, or interference with electrical equipment.
Innovation Solution
A method involving circulating a liquid through a closed-loop with a pump driven by an electric motor, sensing pressure and current draw, and determining a leak based on these parameters, utilizing primary and secondary leak detection logic to ensure accurate and rapid detection and isolation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional level sensors and pressure sensors are used for leak detection, then the system can monitor fluid levels and pressure conditions, but the detection reliability is insufficient and may provide inaccurate readings under certain pump conditions
Solution Approach 1:
The patent combines multiple sensor types (level sensors, pressure sensors, and flow sensors) into an integrated leak detection system that cross-validates readings from each sensor to determine leak presence. This merging of sensing modalities resolves the reliability and accuracy limitations of individual sensors by using their complementary strengths.
Solution Approach 2:
The system continuously monitors sensor readings and uses feedback logic to adjust leak detection thresholds based on actual system operating conditions. The control system compares expected sensor readings under normal operation with actual readings, and dynamically adjusts detection parameters to maintain high reliability and accuracy across varying pump conditions.
2Loss of time
If leak detection sensitivity is increased to detect leaks early, then leaks can be isolated before significant fluid loss, but false alarms may increase and reduce system reliability
Solution Approach 1:
The system uses multiple sensing modalities (level, pressure, and flow sensors) rather than relying on a single sensor type. This partial redundancy allows the system to detect leaks with high sensitivity while cross-validating readings to filter out false alarms, thereby reducing the false alarm rate while maintaining early leak detection capability.
Solution Approach 2:
The leak detection thresholds and sensitivity parameters are dynamically adjusted based on system operating conditions such as pump speed, fluid temperature, and expected flow rates. This allows the system to maintain high detection sensitivity across varying conditions while adapting thresholds to minimize false alarms caused by normal operational variations.
3Measurement precision
If multiple sensors are added to improve leak detection accuracy, then detection reliability improves, but device complexity increases
Solution Approach 1:
The control system serves multiple functions: it manages normal pump operation, processes readings from multiple sensor types, performs leak detection, and controls leak isolation valves. By making the control system universal and multi-functional, the patent avoids adding separate dedicated systems for each function, thereby improving leak detection accuracy without proportionally increasing overall system complexity.
Data Source
Figure 1
Figure 2
AI summary
A method for detecting a leak in a closed-volume liquid system (10) comprises circulating a fluid through a closed-loop with a pump (22) having a reservoir (24) and that is driven by an electric motor (23). The method also comprises sensing a pressure in the closed-loop, sensing current draw by the motor (23) and sensing fluid level in the reservoir (24). The method further comprises determining a presence of a leak of fluid from the closed-loop based upon two of the sensed signals using a first leak detection logic (150), and determining the presence of a leak of fluid from the closed-loop based upon at least one of the sensed signals using a second leak detection logic (150).